CT presentation and staging accuracy of pancreatic adenocarcinoma.

In a retrospective study of 56 patients with surgically confirmed adenocarcinoma of the pancreas, the authors analyzed the efficacy of contrasted CT examination in the staging of tumor resectability and unresectability. Resectability was defined as strictly intrapancreatic disease; unresectability as locally invasive or metastatic disease. Of the 56 patients, 49 (88%) were found to be unresectable, and only seven (12%) were considered potentially resectable on the basis of CT evaluation. When these findings were compared to surgical findings, only one of the 49 CT unresectable patients was considered surgically resectable; all others showed advanced disease, with 30 of 49 (61%) having surgical findings comparable to CT findings, and 18 of 49 (37%) having more extensive disease at surgery than at CT. Of the 7 patients judged resectable by CT, only two (29%) were found to have local disease at surgery; three were locally invasive, one had positive nodes, and one had liver metastases. This study would suggest that CT can be very reliable in distinguishing advanced disease from confined intrapancreatic lesions (unresectability), but is not reliable for predicting locally confined disease (resectability). In particular, CT seems to be ineffective in demonstrating the full extent of pancreatic carcinoma, either unresectable or resectable, because of its inability to demonstrate microscopic local and nodal invasion and surface metastases.